Artificial Intelligence Policy Tracker
Search
Menu

AI incident ·

Grok AI Reportedly Generated Fabricated Civilian Hero Identity During Bondi Beach Shooting

3 news reports Snapshot 7 Sep 2026
On this page 9 sections

In brief

An AI system built and deployed by Xai allegedly harmed General Public Of Sydney, General Public Of Australia and 5 others.

Risk domain
Misinformation False or misleading information
Occurred
Coverage
3 reportsDec 2025

What happened

Grok reportedly generated and repeated a fabricated civilian hero identity, "Edward Crabtree," following the Bondi Beach shooting in Sydney, Australia. The system reportedly cited a fake news article and misattributed heroic actions to this fictional individual during the unfolding emergency. Contemporaneous reporting identified Ahmed Al Ahmed as the real bystander who intervened and was injured during the attack.

Laws on record for this risk area

Policy angle: Classified under Misinformation (False or misleading information) in the MIT AI Risk Repository taxonomy; 5 instruments on record cover this risk area.

Matched from the record's risk category and country to the instruments in effect recorded here: a law listed covers the area, which is not a finding that it applies to this incident. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).

News reports (3)

Titles link to the original publisher; report text is not reproduced here.

  1. Grok Is Glitching and Spewing Misinformation About The Bondi Beach Shooting
    gizmodo.com · Ece Yildirim
  2. The 5 fake Bondi attack stories spread by AI and social media
    afr.com · Sam Buckingham-Jones, Amelia McGuire
  3. Fake hero, wrong suspect: Misinformation floods social media after Bondi shooting
    smh.com.au · David Swan

Who was involved

Alleged deployer
Xai
Alleged developer
Xai
Alleged harmed party
General Public Of Sydney, General Public Of Australia, General Public, Epistemic Integrity, Bondi Beach Shooting Bystanders, Ahmed Al Ahmed, National Security And Intelligence Stakeholders

Classification (MIT AI Risk Repository taxonomy)

Risk domain
Misinformation
Causal entity
AI
Intent
Unintentional
Timing
Post-deployment
Harm level
—
Sectors
—
Countries
—

Risk entries describing this failure mode

Entries from the MIT AI Risk Repository coded to subdomain 3.1.

  • Pursuing Consistent Context

    "LLMs have been demonstrated to pursue consistent context [129]–[132], which may lead to erroneous generation when the prefixes contain false information. Typical examples include sycophancy [129], [130], false demonstra...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Defective Decoding Process

    In general, LLMs employ the Transformer architecture [32] and generate content in an autoregressive manner, where the prediction of the next token is conditioned on the previously generated token sequence. Such a scheme...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Noisy Training Data

    "Another important source of hallucinations is the noise in training data, which introduces errors in the knowledge stored in model parameters [111]–[113]. Generally, the training data inherently harbors misinformation....

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Untruthful Content

    "The LLM-generated content could contain inaccurate information"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Hallucinations

    "LLMs generate nonsensical, untruthful, and factual incorrect content"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Faithfulness Errors

    "The LLM-generated content could contain inaccurate information" which is is not true to the source material or input used

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Factuality Errors

    "The LLM-generated content could contain inaccurate information" which is factually incorrect

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Knowledge Gaps

    "Since the training corpora of LLMs can not contain all possible world knowledge [114]–[119], and it is challenging for LLMs to grasp the long-tail knowledge within their training data [120], [121], LLMs inherently posse...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

Incidents in the same risk subdomain

All incidents in this subdomain

Other incidents involving Xai

Source record: incident #1307 on the AI Incident Database · all 3 reports